Skip to content

The CCS Interview

Braavos CEO: Starknet's built-in account abstraction could end the seed phrase

Motty Lavie, CEO and co-founder of Braavos, explains why his team built a smart contract wallet on Starknet and what account abstraction makes possible. The conversation also covers the ecosystem's growth, wallet security and staking costs. It was recorded in February 2024, and Lavie's plans for 2024 are given as he described them then.

Interview by Ashton Addison Recorded February 2024 6 min read

Key takeaways

  • Starknet uses validity rollups with ZK proofs for scalability while maintaining Ethereum's decentralization and security standards.
  • Account abstraction baked into Starknet's protocol enables custom signature verification and arbitrary execution logic per transaction.
  • Smart contract wallets like Braavos offer two-part architecture with client-side application and on-chain account contract for enhanced functionality.
  • Starknet's cheap computation and account abstraction together address the user experience gap between traditional web applications and self-custodial crypto.
  • Braavos wallet is necessary on Starknet because the protocol requires smart contract wallets, unlike chains where MetaMask can be used directly.

From Web2 to a Starknet wallet

Ashton Addison: What is your background, and what led you to start Braavos?▶ 00:23

Motty Lavie: I'm a latecomer to crypto. I spent most of the last decade in a company I co-founded, a classic Web2 company that did everything from embedded software to cloud software, and a lot of AI before AI was as popular as it is today. About two and a half years ago I had the luxury of taking the time to study the space, and I fell in love with it. But coming from Web2, it was clear there was a huge gap in user experience. There are a lot of good things in crypto, but the experience needs a lot of improvement if we ever want mass adoption. Centralized crypto might be a good entry point, but it's not what crypto is all about. My co-founder and I said this gap has to be closed, and a few things came together at the right time: cheap computation and account abstraction.

Ashton: Why did you choose to build on Starknet?▶ 03:29

Motty: When I studied the different L1s, I quickly concluded that Ethereum made the right trade-off between decentralization, scalability and security. It sacrificed scalability to get decentralization and security. Then I looked at how to scale it. Sidechains clearly weren't a real solution. Optimistic rollups were at best an intermediate solution, because they have a strong limit on how much they can scale and their finality can take up to a week. It was clear that validity rollups, or ZK rollups, were the future. At the time there were two main players, zkSync and Starknet. Starknet was far more advanced, with battle-tested proving technology already working in StarkEx, which protocols like dYdX and Immutable X used. The main concern was that Starknet invented its own language, Cairo, optimized for validity proofs, instead of using Solidity and the EVM. We built a small test project to check that it worked, and it became very successful. Starknet gave us abundant, cheap computation, and it had baked account abstraction into the protocol.

In crypto you can hold millions and have no 2FA, and everything ends up in a 12 or 24 word phrase.

Account abstraction, explained

Ashton: Can you explain account abstraction and why it matters?▶ 07:15

Motty: Account abstraction, which technical listeners may know as ERC-4337, is what sits at the base of smart contract wallets. Traditional wallets like MetaMask are application-only. The application, either a browser extension or a phone app, reflects your account's status on the blockchain and signs transactions, but everything happens on the client side. A smart contract wallet has two sides: the application, like a traditional wallet, and an account smart contract on chain that all your transactions go through. That is extremely powerful because it allows two main things. One is custom signature verification logic, so we're no longer bound to the signature scheme the blockchain set. The other is arbitrary execution logic, so we can add any logic we want to each transaction. That's a whole new design space, and we don't sacrifice decentralization or self-custody.

Ashton: Is that why you can't simply add Starknet to MetaMask the way you can with zkSync or Linea?▶ 09:23

Motty: Exactly. Because Starknet baked account abstraction into the protocol, every wallet on it has to be a smart contract wallet, and the whole ecosystem works with them. Having that from the start is powerful because you don't have the compatibility issues you get on classic EVM chains that rely on externally owned accounts like MetaMask. A classic example is a swap. On Uniswap on Ethereum L1 you first approve the AMM to use your token, usually for the maximum amount, so it has the allowance to withdraw whatever it wants. Only in a second or third transaction do you do the swap. On Starknet, one of the basic features is multicall. You can combine the approval and the swap in one transaction, approve exactly the amount you want to swap, and not give protocols an infinite allowance into your wallet.

Building and growing on Starknet

Ashton: Is Starknet harder for existing dapps like Uniswap to integrate than other EVM chains?▶ 12:20

Motty: It's a blessing and a curse, because the bar is higher. For Uniswap, moving from Ethereum L1 to Arbitrum or Optimism is practically a click of a button, because it's the same Solidity code that is already working, audited and debugged. To onboard to Starknet they would need to rewrite their entire code base from Solidity to Cairo and adjust it to Starknet's functionality. That's heavy lifting, maybe years of engineering work. The upside is that new protocols and approaches can emerge on Starknet without having to fight incumbents. Take Braavos. MetaMask has over 80% market share on Ethereum, and a new startup fighting it faces an extremely hard uphill battle. On Starknet we can build a user base and show the advantage of smart contract wallets without that battle, and show what can be done with account abstraction.

Ashton: How fast is the Starknet ecosystem growing?▶ 15:00

Motty: In terms of developers, I think it's the fastest-growing ecosystem by far. More and more developers are coming to learn Cairo and are trying to build projects in it. We came to Starknet at the very beginning, when there weren't many teams and we knew everyone. Now the ecosystem is not only growing but accelerating, which is really encouraging. Our own initial project was a basic constant-product AMM, built just to see whether we could build something on Starknet. It became so popular that we launched it on mainnet, then upgraded it to a singleton, concentrated-liquidity AMM that uses Starknet's cheap computation to offer better quotes and prices. There are a lot of projects like that. Starknet ID, for example, is experimenting with onchain name servers and profile pictures. With no incumbents, the door is open to a lot of innovation.

We want to get rid of the seed phrase completely, which I think will be a major milestone for the industry.

Security, cheaper staking and the road ahead

Ashton: How do you convince users to choose Braavos as their main Starknet wallet?▶ 17:55

Motty: We approached it from two angles. First, since Starknet is new and there's no Compound or Uniswap yet, people don't know what to experiment with. So we built the Starknet Journey, which monitors your interactions with each dapp and builds a gamified proficiency score. People really liked it. Second, there are the core elements of the product. There are gazillions of wallets, but not gazillions of smart contract wallets, and we constantly try to explore uncharted territory. We built a truly decentralized two-factor authentication. It exists in every mediocre Web2 service, even when I order food, but in crypto you can hold millions with no 2FA, and everything ends up in a 12 or 24 word phrase. Recently we added three-factor authentication with a passkey you can use from the browser. We also democratized staking ETH. On L1, using an aggregator like Lido or Rocket Pool can cost $10, $20 or $30 to deposit and about the same to withdraw, which makes small investments uneconomical. Because computation is cheap on Starknet, we pool users' funds, ship them to Ethereum L1 as one transaction, stake in Lido, bridge the staked token back to Starknet and distribute it to users. Instead of a $20 fee, you pay 20 or 30 cents.

Ashton: What are your plans for 2024?▶ 22:37

Motty: We want to close the gap between Web2 and Web3. That means bringing a Web2 experience with Web3 values, without compromising decentralization or self-custody. For that we still have a lot to do. We want to get rid of the seed phrase completely, which I think will be a major milestone for the industry. Without sacrificing decentralization or self-custody, we'd have no seed phrase and still have 2FA. That would help users who join this cycle graduate from centralized exchanges like Coinbase and Kraken, if they're lucky, or Celsius and FTX, if they're less lucky, into real self-custodial crypto. Self-custodial solutions are still extremely complex and unfriendly, and our goal is to close that gap.

Ashton: You recently ran a public security challenge. What happened?▶ 25:18

Motty: Last week we ran a Capture the Flag initiative. We put 150,000 USDC in a wallet and gave the seed phrase to the entire Twitter community to see if anyone could hack it. Numerous people recovered the wallet in Braavos and saw the high balance on their app or computer, but couldn't do anything because they didn't have the 2FA. For many people that was quite mind-boggling.

This interview has been edited for length and clarity. Recorded February 2024.

Motty Lavie
Motty LavieMotty Lavie is the CEO and co-founder of Braavos, a self-custodial smart contract wallet for Starknet and Bitcoin, based in Tel Aviv. He co-founded Braavos with Abraham Makovetsky and Yoav Gaziel, and the company was founded in early 2022. Before Braavos he was co-founder and… Full profile →
Share this interviewXLinkedIn
▶ Watching the interview